The incomplete removal of time-varying effects in the Magsat data leads to the necessity of fitting low-order polynomial functions in order to minimize differences between passes. This process makes the estimation of the zero level in the crustal anomaly field very difficult. Refinements in main field analysis, ring current estimation, and the identification of other time-varying field sources should provide us with a physical realization of these differences. The spectral overlap of core and crustal anomaly fields makes it particularly difficult to interpret the longer wavelengths of crustal anomalies and refined filtering techniques may need to be developed to improve their resolution. Equivalent source techniques for the inversion of the satellite data to a constant thickness, varying-magnetization source layer solution are well developed. Magnetization and reduced-to-the-pole anomaly maps are now available for many regions. Problems of instability of mathematical inversions are still present and are particularly noticeable near the magnetic equator. The equivalent source solutions contain the same lack of information regarding zero level as the data from which they are derived. Forward modeling techniques are available to model discrete source distributions and are useful in testing among various interpretive hypotheses. Forward modeling techniques are inappropriate for complex geological situations, and some hybrid techniques are envisaged, in which the inverted magnetization are constrained by some a priori estimates of their distribution and variability in order to provide geologically meaningful solutions. There is now sufficient information on the Curie points and magnetic properties of rocks which have been derived from the deep crust and upper mantle to begin to develop magnetization-depth models for a number of tectonic environments. While the familiar Fe-Ti oxides are the magnetic minerals characteristic of the crust, relatively nonmagnetic chromium spinels and magnesian ilmenites are characteristic of the subcontinental mantle; thus the Moho is the magnetic bottom of the continents except where the Curie isotherm is elevated within the crust. Intensity of magnetization in the continental crust is strongly dependent on metamorphic grade and, in general, increases with basicity: mafic granulites tend to form the most magnetic laterally extensive zones of the crust. Some large-scale mafic plutonic complexes also have very large magnetizations. Model studies and laboratory measurements are in agreement that the more magnetic rocks of the crust typically have magnetization values in the range 2--6 A/m. In the deep crust, Curie points are predominantly near that of magnetite, although in hot, anhydrous environments in active tectonic zones, Curie temperatures may be reduced below 300¿C. A comprehensive laboratory program to characterize further the magnetic properties of all lithologies important in the crust is in progress, and much work remains. The goal is an understanding of how crustal evolutionary processes have left their imprint in the magntic mineralogies, and how this in turn is reflected in the anomaly field.